Compare XPUs

Select up to 5 XPUs to compare side-by-side

Select XPUs to Compare

Clear all (4)Jump to results

Filter by Vendor

Showing 72 XPUs • 4 selected

Alibaba

Hanguang 800

AMD

MI100

23.1 TFLOPs

AMD

MI210

181 TFLOPs

AMD

MI250X

383 TFLOPs

AMD

MI300A

980.6 TFLOPs

AMD

MI300X

1,307 TFLOPs

AMD

MI325X

1,400 TFLOPs

AMD

MI350X

2,100 TFLOPs

AMD

Radeon PRO W7900

122 TFLOPs

AMD

Radeon RX 7900 XT

104 TFLOPs

AMD

Radeon RX 7900 XTX

122 TFLOPs

AWS

Inferentia2

190 TFLOPs

AWS

Trainium

190 TFLOPs

AWS

Trainium2

680 TFLOPs

Baidu

Kunlun II

Biren Technology

BR100

Cambricon

MLU370

256 TFLOPs

Cerebras

WSE-3

Enflame Technology

CloudBlazer T20

Etched

Sohu

10,000 TFLOPs

FuriosaAI

Warboy

Google

TPU v4

275 TFLOPs

Google

TPU v5e

197 TFLOPs

Google

TPU v5p

459 TFLOPs

Google

TPU v6e (Trillium)

918 TFLOPs

Graphcore

Bow IPU

Graphcore

IPU-M2000

Groq

LPU Inference Engine

Huawei

Ascend 910B

Iluvatar CoreX

BI-V150

300 TFLOPs

Intel

Data Center GPU Max 1100

177 TFLOPs

Intel

Data Center GPU Max 1550

419 TFLOPs

Intel Habana

Gaudi 2

432 TFLOPs

Intel Habana

Gaudi 3

1,835 TFLOPs

Multi-Metric Comparison

Relative performance across 5 key metrics (normalized to 100 = best in comparison)

Compute Performance (BF16)

Memory Capacity

Power Consumption

Power Efficiency

Specifications

Specification	AMD MI100	Google TPU v4	NVIDIA GB200 NVL72	Qualcomm Cloud AI 100
Architecture	CDNA	TPU v4	Blackwell	Qualcomm AI
Form Factor	PCIe	Mezzanine	Custom	PCIe
VRAM	32 GB	—	13824 GB	16 GB
Memory Bandwidth	1,228 GB/s	—	130,000 GB/s	134 GB/s
TFLOPs (FP32)	23.1	—	6,480	50
TFLOPs (FP16)	184.6	—	360,000	400
TFLOPs	23.1	275	360,000	50
TFLOPs (FP8)	—	—	—	—
TDP	300 W	300 W	120000 W	75 W
Launch Date	Nov 2020	May 2021	Mar 2024	Sep 2020

Efficiency Metrics

Metric	MI100	TPU v4	GB200 NVL72	Cloud AI 100
TFLOPs per Watt (FP32-eq)	0.08	0.46	1.50	0.67
Memory Bandwidth per GB	38.4 GB/s	—	9.4 GB/s	8.4 GB/s

Performance Equivalence

How many units of each GPU are needed to match the performance of the others?

To match 1x AMD MI100

Google TPU v4

Compute (FP32-eq)

0.17x

TPU v4 is 5.95x faster

NVIDIA GB200 NVL72

Compute (FP32-eq)

0.00x

GB200 NVL72 is 7792.21x faster

FP32 Compute

0.00x

GB200 NVL72 is 280.52x faster

VRAM

0.00x

GB200 NVL72 has 432.00x more

Memory Bandwidth

0.01x

GB200 NVL72 has 105.86x more

Qualcomm Cloud AI 100

Compute (FP32-eq)

0.46x

Cloud AI 100 is 2.16x faster

FP32 Compute

0.46x

Cloud AI 100 is 2.16x faster

VRAM

2.00x

Need 2.00x Cloud AI 100

Memory Bandwidth

9.16x

Need 9.16x Cloud AI 100

To match 1x Google TPU v4

AMD MI100

Compute (FP32-eq)

5.95x

Need 5.95x MI100

NVIDIA GB200 NVL72

Compute (FP32-eq)

0.00x

GB200 NVL72 is 1309.09x faster

Qualcomm Cloud AI 100

Compute (FP32-eq)

2.75x

Need 2.75x Cloud AI 100

To match 1x NVIDIA GB200 NVL72

AMD MI100

Compute (FP32-eq)

7792.21x

Need 7792.21x MI100

FP32 Compute

280.52x

Need 280.52x MI100

VRAM

432.00x

Need 432.00x MI100

Memory Bandwidth

105.86x

Need 105.86x MI100

Google TPU v4

Compute (FP32-eq)

1309.09x

Need 1309.09x TPU v4

Qualcomm Cloud AI 100

Compute (FP32-eq)

3600.00x

Need 3600.00x Cloud AI 100

FP32 Compute

129.60x

Need 129.60x Cloud AI 100

VRAM

864.00x

Need 864.00x Cloud AI 100

Memory Bandwidth

970.15x

Need 970.15x Cloud AI 100

To match 1x Qualcomm Cloud AI 100

AMD MI100

Compute (FP32-eq)

2.16x

Need 2.16x MI100

FP32 Compute

2.16x

Need 2.16x MI100

VRAM

0.50x

MI100 has 2.00x more

Memory Bandwidth

0.11x

MI100 has 9.16x more

Google TPU v4

Compute (FP32-eq)

0.36x

TPU v4 is 2.75x faster

NVIDIA GB200 NVL72

Compute (FP32-eq)

0.00x

GB200 NVL72 is 3600.00x faster

FP32 Compute

0.01x

GB200 NVL72 is 129.60x faster

VRAM

0.00x

GB200 NVL72 has 864.00x more

Memory Bandwidth

0.00x

GB200 NVL72 has 970.15x more

Pricing

Price Type	MI100	TPU v4	GB200 NVL72	Cloud AI 100
CAPEX (Street Price)	—	—	—	—
OPEX (per hour)	—	$3.00/hr	—	—
Price per TFLOPs (FP32-eq)	—	—	—	—